JP5821363B2 - Product defect factor analyzer - Google Patents

Description

  The present invention relates to a product defect factor analyzer used in industrial plants such as steel and non-ferrous metals.

  In industrial plants such as steel and non-ferrous metals, various types of plant data are collected and accumulated. When a defect occurs in a product, the accumulated plant data is used to analyze the cause of the defect. .

  There are a great many items that can be collected in plant data. Conventionally, items necessary for analyzing defect factors are selected in advance from plant data, and only the plant data of the selected items is collected and accumulated. In addition, when a defect occurs in a product, a specific item to be investigated is further selected from the accumulated plant data, and the cause of the defect is analyzed using the plant data of the selected item. It was.

  For example, Japanese Patent Application Laid-Open No. H10-228707 proposes a method for automatically calculating a combination having a high degree of influence on product quality from plant data collected and accumulated in advance.

JP 2008-146621 A

  In conventional apparatuses including those described in Patent Document 1, items that may cause product defects are selected in advance from plant data, and only plant data of the selected items is collected and accumulated. In addition, when a defect occurs in a product, only those that are likely to be the cause of the defect are selected for investigation from the accumulated plant data. For this reason, knowing about plant operation, process control, and equipment maintenance is necessary to select the data items to be accumulated and to select the data items to be investigated. Depending on the skill of the user, the cause of the defect can be analyzed correctly. There were cases where it was not possible.

  The present invention has been made to solve the above-described problems, and its purpose is to provide useful information that can cause product defects without requiring know-how regarding plant operation, process control, and equipment maintenance. It is an object of the present invention to provide a product defect factor analyzer that can be easily grasped.

The product defect factor analysis apparatus according to the present invention is a method for analyzing a defect factor of a product by a user in a steel or non-ferrous metal rolling line that manufactures a predetermined product using a plurality of devices connected to a predetermined communication network. Product defect factor analysis device used to perform a time-series collection of all plant data that can be collected from each device via a communication network, and plant data collected by the data collection device. A data storage device to be stored, a product defect detection device for detecting defect data corresponding to a product defect from plant data stored in the data storage device, and a product defect data detected by the product defect detection device. Similar to identifying operating conditions and extracting multiple other products manufactured under operating conditions similar to the operating conditions of the defective product as similar operating products Compare all plant data related to defect occurrence products accumulated in the industry extraction device and data accumulation device with corresponding plant data of similar operation products, and extract abnormal data including abnormal values of plant data of defect occurrence products comprising a data comparison device, to create a display list of defective factor candidates based on the abnormal data extracted by the data comparator, the defect factor candidate editing apparatus for displaying on a predetermined display device, a data comparison device, the defect For the generated product and the similar operation product, first, the plant data at a predetermined first time interval is compared. After that, if the comparison is not completed for all the plant data, the first time interval is set for the uncompared plant data. The plant data of the second time interval shorter than the second time interval are compared .

  With the product defect factor analysis device according to the present invention, useful information that can cause product defects can be easily grasped without requiring know-how regarding plant operation, process control, and facility maintenance. .

It is a figure which shows the structure of the product defect factor analyzer in Embodiment 1 of this invention. It is the equipment schematic of the industrial plant to which the product defect factor analyzer shown in FIG. 1 is applied.

In order to explain the present invention in more detail, it will be described with reference to the accompanying drawings .

Embodiment 1 FIG.
FIG. 1 is a diagram showing a configuration of a product defect factor analyzer according to Embodiment 1 of the present invention. This product defect factor analysis device analyzes product defect factors in an industrial plant that manufactures a predetermined product using a plurality of devices connected to a predetermined communication network, for example, an industrial plant such as steel or non-ferrous metal. Used to do.

The industrial plant includes a process computer 1, a process controller 2, a drive means (electric motor / drive device, etc.) 3, a process sensor 4, and an operator operation terminal 5.
The analysis device 6 constituting the main part of the present invention is connected to the process computer 1 and the operator operation terminal 5 via a predetermined information LAN 7. The analysis device 6 is connected to the process computer 1, the process controller 2, the driving means 3, and the process sensor 4 via a predetermined control LAN 8.

FIG. 2 is an equipment schematic diagram of an industrial plant to which the product defect factor analyzer shown in FIG. 1 is applied. Specifically, FIG. 2 shows a hot rolling line for steel. In the steel hot rolling line, the slab extracted from one or a plurality of heating furnaces 9 undergoes rough rolling by a roughing mill 10, finish rolling by a finishing mill 11, and cooling by a run-out table (ROT) 12. It is wound up as a coil in the winder 13. The wound coil is bundled, weighed, etc., and then conveyed to the downstream process by the coil conveyor 14.

  The analysis device 6 is for collecting, accumulating and processing plant data covering a number of items from each device constituting an industrial plant (for example, a hot rolling line shown in FIG. 2). The analysis device 6 includes a data collection device 15, a data storage device 16, a product defect detection device 17, a similar operation extraction device 18, a data comparison device 19, a defect factor candidate editing device 20, a storage device 21, and a display terminal 22. ing.

  The data collection device 15 has a function of collecting plant data from each device used in an industrial plant. As described above, each device in the industrial plant is connected to a predetermined communication network (in this embodiment, the information LAN 7 and the control LAN 8). The data collection device 15 collects all plant data that can be collected via the communication network in time series.

  The information LAN 7 is an information network related to operations and product manufacturing. The data collection device 15 collects product manufacturing instructions, control amount settings for each device, product quality results, and the like from the process computer 1 as plant data for each event (operation event) via the information LAN 7. Further, the data collection device 15 collects operation history as plant data from the operator operation terminal 5 via the information LAN 7.

The data collection device 15 collects real-time data as plant data from the process controller 2, the drive unit 3, and the process sensor 4 via the control LAN 8.
The control LAN 8 is a process control network that provides a data area shared between nodes. The high-speed data area of the control LAN 8 enables data transmission with a cycle of 2 [ms], for example. The control output of the process controller 2, the control data of the driving means 3, and the measurement output value of the process sensor 4 are output to the shared data area of the control LAN 8. The data collection device 15 collects all data in the shared data area of the control LAN 8 in accordance with the transmission cycle.

  The data collection device 15 is composed of one or a plurality of units. Since the number of items and the speed of the collected plant data differ depending on the scale of the industrial plant, the data collection device 15 is configured according to the installation environment.

The data storage device 16 includes a large-capacity database that stores plant data collected by the data collection device 15. In the data storage device 16, various plant data collected by the data collection device 15, for example, control information of each facility, event information, operation history information, alarm information, operation information such as product manufacturing instructions and product results, Accumulated for a predetermined period in time series. The capacity required for the data storage device 16 is determined by, for example, the number of items of plant data to be stored, the collection cycle, and the storage period. The period during which the data storage device 16 stores the plant data takes into consideration the accuracy of extraction by the similar operation extraction device 18 described later, for example, about 3 months to half a year in the steel hot rolling line shown in FIG. .
In addition, the data storage device 16 has a function of searching and providing stored data in response to a request from another device.

  The product defect detection device 17 has a function of detecting that a specific defect has occurred in the product. The product defect detection device 17 detects defect data corresponding to a product defect from the plant data stored in the data storage device 16. For example, in the steel hot rolling line shown in Fig. 2, product defects (product thickness, plate width, temperature, shape, etc.) deviate from target values or nominal values, Or deformation. When a specific event (for example, a predetermined information input indicating coil winding completion, wrinkles or deformation from the operator operation terminal 5) occurs, the product defect detection device 17 extracts the actual data of the manufactured product. Investigate the presence of product defects (defect data). When the product defect detection device 17 detects the defect data, the product defect detection device 17 outputs information on the product in which the defect has occurred and information on the defect to the similar operation extraction device 18 as product defect information.

  The similar operation extraction device 18 has a function of extracting a product manufactured by the same operation as when a defect occurs in the product. When the similar operation extraction device 18 receives the product defect information from the product defect detection device 17, the similar operation extraction device 18 specifies the operation condition of the product in which the defect data is detected by the product defect detection device 17 (hereinafter also referred to as “defect occurrence product”). . Next, based on the plant data stored in the data storage device 16, the similar operation extraction device 18 uses another product (hereinafter referred to as “similarity”) manufactured under the operation condition similar to the operation condition of the identified defect occurrence product. Extraction of “operational product”. The number of products to be extracted as similar operation products is set in advance to a predetermined number (for example, 5).

  Specifically, the similar operation extraction device 18 extracts products whose material specifications, product specifications, control target values, and equipment used during manufacture all match the defect-occurring products as similar operation products. In the hot rolling line for steel shown in FIG. 2, the similar operation extraction device 18 includes specifications of each slab (material) and coil (product) (steel type, plate thickness, plate width, weight, etc.), control target value (extraction). Temperature, coiling temperature, target values such as sheet thickness and sheet width at the completion of rough rolling) are the same as those of the defective product, and the heating furnace 9 and the winder 13 used are defective products. Extract the same products as the ones.

When the number of similar operation products extracted under the similar condition (condition for regarding the operation condition as similar) does not reach the set number, the similar operation extraction device 18 relaxes the similar condition and the similar operation product. Perform the extraction again. For example, the similar operation extraction device 18 extracts, as similar operation products, products whose material and product specifications match those of the defective product and whose equipment is different from that of the defective product. In addition, when the number of similar operation products does not reach the set number due to the similar conditions after relaxation, the similar operation extraction device 18 further relaxes the extraction conditions regarding the product specifications and the control target value, A product whose control target value falls within a predetermined range is extracted as a similar operation product.
When the similar operation extraction device 18 extracts the set number of products (or less than that number) as similar operation products, the similar operation extraction device 18 outputs each piece of information on the defect occurrence product and the similar operation product to the data comparison device 19.

  The data comparison device 19 has a function of comparing the plant data of the defective product with the plant data of the similar operation product. When the data comparison device 19 receives the information from the similar operation extraction device 18, the data comparison device 19 compares all the plant data related to the defect occurrence product stored in the data storage device 16 with the corresponding plant data of the similar operation product.

  Specifically, the data comparison device 19 first extracts the occurrence time of a predetermined event. In the hot rolling line for steel shown in FIG. 2, the event starts when the extraction of the slab from the heating furnace 9 is started (extraction start). The start and completion of rough rolling by the rough rolling mill 10, the start and completion of finish rolling by the finish rolling mill 11, and the start and completion of coil winding by the winder 13 are also events. Information about each event is stored in the data storage device 16.

  When the event comparison time is extracted, the data comparison device 19 performs time correction for matching the event intervals between the defect occurrence product and the similar operation product. By this time correction, the data comparison device 19 can compare the plant data having the same collection timing with respect to the event period for the defect occurrence product and the similar operation product. For example, if it takes 30 seconds for a defective product and 35 seconds for one similar operation product from the start of slab extraction by the heating furnace 9 to the start of rough rolling by the roughing mill 10, 10 seconds from the start of slab extraction The plant data of the later defective product is compared with the plant data of the similar operation product after 10 × 35/30 seconds.

  When the data comparison device 19 performs the time correction, the data comparison device 19 compares the plant data obtained at predetermined time intervals (for example, every 5 seconds). For example, if it takes 300 seconds (time after time correction: coincides between the defective product and each similar operation product) from the start of slab extraction by the heating furnace 9 to the completion of winding by the winder 13, the first time If the interval is set to 5 seconds, the data comparison device 19 first compares plant data for 300/5 = 60 sets.

  When all the comparisons of the 60 sets of plant data are completed, the data comparison device 19 sets the time interval to a time interval shorter than 5 seconds (for example, 1 second) and compares the plant data. At this time, since the comparison is already completed for the plant data at intervals of 5 seconds, the comparison of the plant data is performed only for those that have not been completed yet. That is, when the comparison for the 60 sets is completed, the data comparison device 19 then compares the plant data for 300 / 1-60 = 240 sets.

  Thereafter, the data comparison device 19 gradually shortens the time interval for comparison to 200, 50, 10, 2 [ms], and performs the same processing as above until the comparison of all plant data is completed. repeat.

  When comparing plant data, for example, the data comparison device 19 determines whether or not the value of the plant data of the defective product is out of the value of the corresponding plant data of the similar operation product. That is, when the value of the plant data of the defective product is out of the range where the maximum value of the corresponding plant data corresponding to the similar operation product is the maximum and the minimum value is the minimum, the data comparison device 19 Are detected as abnormal values that are candidates for product defect factors.

  Note that, as described above, if an abnormal value is detected while the time interval for comparison is gradually narrowed, an abnormal value that can be detected even by comparing coarse time intervals can be detected at an early stage. Finally, all the plant data can be compared by shortening the time interval until the transmission period of the control LAN 8.

  When the data comparison device 19 detects an abnormal value from the plant data of the defective product, the data comparison device 19 extracts the abnormal data based on the plant data stored in the data storage device 16. In the abnormal data, in addition to the data items, values, and times of those detected as abnormal values, the position of the target material, the position of the preceding material and the succeeding material, the average value of the plant data corresponding to the abnormal value of similar operation products・ The maximum and minimum values are included. When the data comparison device 19 extracts the abnormal data, the data comparison device 19 outputs the information to the defect factor candidate editing device 20.

  In addition, like the heating furnace 9 and the winder 13 of the hot rolling line shown in FIG. 2, when there are a plurality of devices of the same type and different devices are used depending on the product, the data comparison device 19 compares them. You may replace the plant data of similar operation products with the equipment used at the time of manufacture of a defect generation product.

  The defect factor candidate editing apparatus 20 has a function of creating a display list of defect factor candidates and displaying the defect factor candidate display list on the display terminal 22. When the defect factor candidate editing device 20 receives the abnormal data from the data comparison device 19, the defect factor candidate editing device 20 creates a display list based on the received abnormal data.

  In the storage device 21, items of which plant data stored in the data storage device 16 is premised on different values for each product are stored in advance as excluded data item definitions. Excluded data items include, for example, material and product identification numbers, time information such as slab extraction completion time, and electricity / water / gas consumption integrated values. The defect factor candidate editing device 20 displays predetermined items that do not need to be displayed on the display terminal 22 based on the excluded data item definition stored in the storage device 21 when the defect factor candidate display list is created. Exclude from the list.

  Further, the defect factor candidate editing device 20 determines the item order of plant data to be displayed on the display terminal 22 based on the abnormality data received from the data comparison device 19. For example, the defect factor candidate editing device 20 determines each abnormal data based on the position where the abnormal value is generated (the position of the target material) and the position of the preceding material and / or the succeeding material when the abnormal value occurs. Weighting is performed. This weighting is a position in the display list. That is, when the weight is large, information related to the abnormal value is arranged at the top of the display list, and the display priority on the display terminal 22 becomes high. On the contrary, when the weight is small, the information related to the abnormal value is arranged at the lower position of the display list, and the display priority on the display terminal 22 is lowered. For example, the defect factor candidate editing device 20 sets the weight to be smaller as the position where the abnormal value is generated is farther from the position of the target material. Further, the defect factor candidate editing device 20 sets the weight smaller as the position where the abnormal value is generated and the position of the preceding material or the succeeding material are closer to each other.

The abnormal data received by the defect factor candidate editing device 20 from the data comparison device 19 is data at a specific time, and does not become time-series data because the time fluctuates. The defect factor candidate editing device 20 edits each abnormal data into a series of data based on the item and time, and rearranges them in the order based on the weighting to create a display list. The display list includes, for example, data items, average / maximum / minimum values of abnormal values, abnormal start time, abnormal end time, and average / maximum / minimum values of plant data corresponding to the abnormal values of similar operation products. Contains the value.
When the defect factor candidate editing device 20 creates the display list, the defect factor candidate editing device 20 transmits the data (display data) to the display terminal 22.

  When the data comparison device 19 detects an abnormal value while gradually reducing the time interval for comparison, the defect factor candidate editing device 20 detects an abnormal value (extracts abnormal data). ) New abnormal data is received from the data comparison device 19 each time. In such a case, when the defect factor candidate editing device 20 receives new abnormal data from the data comparison device 19, the defect factor candidate editing device 20 updates the display list once created based on the received information, and transmits the updated display data to the display terminal 22. .

The display terminal 22 displays the display list created by the defect factor candidate editing device 20 on the screen based on the display data received from the defect factor candidate editing device 20.
Note that the display terminal 22 may have a function of displaying a product defect occurrence history and specifying a product defect to be analyzed, and inputting a start and interruption of a series of analysis processes related to the specified product defect.

  If it is a product defect factor analyzer having the above configuration, even a user who does not have sufficient know-how regarding plant operation, process control, and facility maintenance may cause product defects from the display content of the display terminal 22. It is possible to easily grasp useful information to be obtained. Further, by using this product defect factor analyzer, plant data that has not been considered as a factor of product defects in the past can be used for the analysis of defect factors.

DESCRIPTION OF SYMBOLS 1 Process computer 2 Process controller 3 Drive means 4 Process sensor 5 Operator operation terminal 6 Analyzer 7 Information LAN
8 Control LAN
DESCRIPTION OF SYMBOLS 9 Heating furnace 10 Rough rolling mill 11 Finish rolling mill 12 Run-out table 13 Winding machine 14 Coil conveyor 15 Data collection device 16 Data storage device 17 Product defect detection device 18 Similar operation extraction device 19 Data comparison device 20 Defect factor candidate edit device 21 Storage device 22 Display terminal

Claims (7)

A product defect factor analyzer that is used by users to analyze product defect factors in a steel or non-ferrous metal rolling line that manufactures a given product using a plurality of devices connected to a given communication network. There,
A data collection device that collects all plant data that can be collected from each device via the communication network in time series,
A data storage device for storing plant data collected by the data collection device;
A product defect detection device for detecting defect data corresponding to a product defect from the plant data stored in the data storage device;
Similarity is to identify the operation condition of the product for which defect data is detected by the product defect detection device, and to extract a plurality of other products manufactured under the operation condition similar to the operation condition of the defect occurrence product as similar operation products An operation extraction device;
Data for comparing all plant data related to the defect-producing product stored in the data storage device with corresponding plant data of the similar operation product, and extracting abnormal data including abnormal values of the plant data of the defect-generating product A comparison device;
A defect factor candidate editing device for creating a display list of defect factor candidates based on the abnormal data extracted by the data comparison device and displaying the defect factor candidate on a predetermined display terminal;
Equipped with a,
The data comparison device first compares plant data at a predetermined first time interval with respect to the defect occurrence product and the similar operation product, and then compares all the plant data if comparison has not been completed. A product defect factor analyzer for comparing plant data of a second time interval shorter than the first time interval with respect to non-plant data .   2. The product defect factor analysis device according to claim 1, wherein the similar operation extraction device extracts similar operation products based on material specifications, product specifications, control target values, and equipment used during manufacturing.   The data comparison device detects an abnormal value when the value of the plant data of the defective product is out of a range in which the maximum value of the corresponding plant data of the similar operation product is maximum and the minimum value is minimum. The product defect factor analyzer according to claim 1 or 2.   When comparing the plant data between predetermined events for the defect occurrence product and the similar operation product, the data comparison device performs time correction for matching the event intervals, and the collection timing for the period coincides. The product defect factor analyzer according to any one of claims 1 to 3, wherein plant data are compared with each other. The defect factor candidate editing device updates the display list when new abnormal data is extracted by the data comparison device after creating a display list of defect factor candidates, and displays the updated contents on the display terminal. The product defect factor analyzer according to claim 1 . Among the plant data stored in the data storage device, a storage device that stores definitions of items that are premised on different values for each product;
Further comprising
The defect factor candidate editing apparatus based on the stored items defined in the storage device, according to claims 1 to create a display list by excluding certain items of plant data to any one of claims 5 Product defect factor analysis device. The defect factor candidate editing device determines the order of plant data items to be displayed on the display terminal based on the occurrence position of the abnormal value and the position of the preceding material or the position of the succeeding material when the abnormal value occurs. The product defect factor analysis device according to any one of claims 1 to 6 , wherein the defect factor candidate display list is determined and created.

Images